I think this pretty succinctly sums up my thoughts on this matter.
The way some people are arguing in this thread, you’d think Win Expectancy actively interfered with enjoyment of the game, actually detracts from our understanding somehow.
First, I’m well aware that weather forecasts are more sophisticated than that. Second, the WE stat for baseball is more sophisticated than you seem to appreciate, even given its historical nature. There have been over 200,000 major league baseball games in the last century and a half. That’s over three and a half million half-innings, over ten million outs, and millions more plate appearances. That means that the WE stat is based on a lot of data, and as your dataset gets bigger for things like this, it gets more and more likely that the large sample sizes will, in some measure at least, smooth over and hide statistical aberrations and give you a reasonably accurate predictive number.
Does that mean that WE is going to accurately predict the outcome of every game? Of course not, just like an 80% chance of rain doesn’t always mean that you’ll need your umbrella. But it gives a good sense of likely outcomes based on historical data. And there’s nothing to stop you, in your own mind, from supplementing that number with what you know about the peculiar situation that you’re watching. If you know that the reliever who’s just been brought in to pitch has been struggling over the past few weeks, and he’s just about to go up against the heart of a batting order that has been crushing it this season, you might well anticipate that the hitting team will outperform its win expectancy. But that doesn’t mean that WE is useless.
Like RickJay said, I just don’t understand why you’d want less information rather than more, when there’s no trade-off and you lose nothing.
How exactly is WP “more sophisticated than (I) seem to appreciate”)?
Hard to believe you’re still defending that goofy weather forecast analogy, but apparently some baseball fans go into a panic over any perceived threat to a statistic, no matter how useless it is.
I disagree with you, so I must be in “a panic”? Sure, whatever.
Remind me again, who was it who devoted a post to whining about the stat in the first place? Maybe it would be more accurate to say that some baseball fans go into a panic over stats that they don’t like, and could quite easily ignore with little or no effort.
Hey babe, all I said was that I found it a useless stat. From your and RickJay’s reaction, it was as if I had challenged the Sacred Institution of Baseball Sabermetrics.
Don’t worry, there’s nothing stopping you from accessing box scores on an inning-by-inning basis to see how the historical odds of winning/losing are shifting for your favorite team.
Meh. I like the stat, too. It does help to put into perspective how big a comeback was, or just how awful your team was for blowing that lead, and just looking at a graph of win probability with no further context can visually show you at a glance how “exciting” a game was. It’s easy enough to ignore if you don’t like it. And its predictive power should be statistically analyzable.
This statement is directly contradicted by evidence from your own post. After saying that you found it to be useless, you went on to say:
And yet this is literally untrue. It is not meaningless. It means exactly what it says it means; no more, no less. You also said:
Yeah, that’s what the stat people mean when they talk about WinEx - that you don’t need to play the rest of the game at all.
If all you had said was that you find the stat uninteresting or useless for you and your enjoyment of baseball, we probably wouldn’t be having this conversation. It’s the other bullshit, the misrepresentation of what it purports to do, and what its significance is, that prompted my response.
Speaking of bullshit: our weather forecast today calls for mostly cloudy skies and a 20% chance of thunderstorms.
I should contact the National Weather Service to see if they based that forecast on historical averages for this date since 1890 (which you apparently believe), or if they used advanced metrics like approaching fronts, satellite imagery and so on.
Just chill, no bad men are going to take away your cherished WP stat any time soon.
Nowhere have I even once said that weather forecasts are based on historical averages determined by date. I challenge you to find where I’ve even implied it.
Weather forecasts use historical data, but it’s not simply the date as you so absurdly impute to me here. They use historical data about the very things that you mentioned, like fronts, satellite imagery, and other actual weather developments. And one part of their forecasting model is to study what sort of weather patterns have historically emerged under circumstances similar to those currently observed. Part of the weather forecasters’ ability to understand what that front is likely to do, or what that satellite imagery tells us about future weather patterns, is precisely based on their scientific and data-driven understanding of what these types of weather patterns have done in the past.
Please don’t misrepresent my arguments like this; it doesn’t reflect well on you.
I don’t find it meaningless at all. If you want to make in-game bets, it’s a pretty good metric for figuring out what odds to give somebody to make a fair bet. I mean, it’s not saying anything more than that, just like 538’s probability models on political races. For me, it’s fun to see who beats historical expectations, who doesn’t, and how much against the odds they went to mount a comeback. You obviously don’t see those numbers as meaningful – I do. You can come up with a Brier score if you want (I’m not sure if anyone has) to check the validity of the current baseball win probability models. These are models that can be mathematically validated or rejected, not some silly pseudo-science that is immeasurable.
WP is just an estimate of how likely a team is to win given the current game situation. Any fan of any sport thinks about this constantly. All WP does is put into numbers the thing that fans are usually most interested in.
WP does something else. It provides the basis for Win Probability Added (WPA), which is the sum of all the changes in WP for a given player. Like Win Shares, WAR, and several other stats, it gives one more insight into how much a player contributes to his team.
What? Any fan thinks about this constantly? That seems to me to be an extraordinary claim. Fans aren’t actually focused on what the players are actually manage to accomplish on the field only how the probability of the outcome is changing from moment to moment? Why not just make all games into a computer simulation then and just export a series of probabilities?
You may not be aware of this - but many people are able to think about more than 1 thing at the same time. They can both watch a baseball game to enjoy the actual action on the field, AND look at the score/situation and think “I wonder who’ll win this game?”. And many of THOSE people can then do a third thing at that moment, like eat peanuts or drink a beer. It’s amazing.
They’re constantly thinking about the probability of something happening that will definitively revealed in an hour or two?
I can’t imagine why anyone would constantly do that. What does that wondering do for you?
Furthermore, this is a game. The probability of something happening only matters if the game play is based on probabilities. Which is why I said before, it only conceivably matters if what you care about is placing bets in the moment. Pulykamell seems to acknowledge that.
I don’t know if I’m some kind of outlier on this, but it seems to me that being interested in betting and being interested in sports are two discrete interests. So, I’d dispute the claim that “any sports fan is constantly wondering this.” Rather, it would seem more accurate to say “any gambling fan is constantly wondering this.”
I don’t gamble at all, yet I find the win probability fascinating, both during a game and in retrospect. It gives me useful context to look back and see just how important that hit or that strikeout was at the time.
Your arguments against it would seem to apply to anything related to probabilities: if it doesn’t tell you with certainty what will happen, then what’s the point? Obviously a lot of people see the point in applications within baseball and without.
I make about 5 bets a year - so I’m not worried about the outcome of the game for any financial reason.
You don’t watch a game and think how any particular play is going to affect the outcome? That’s the main purpose of watching the game, rather than just checking the scores in the paper the next morning.
And no - I’m not constantly and actively doing that. But I’m most certainly subconsciously doing that. It’s why I recoil when the opposing team hits a home run or my team commits an error, it’s why I get excited when my team hits a triple - because it’s a positive or negative towards my desired outcome.
Do you actually think what the players accomplish on the field is disconnected from the probability of winning?
When a player on my team hits a home run, I’m happy because it means my team is more likely to win. When a pitcher on my team strikes someone out, I’m happy because it means my team is more likely to win. Same thing when my team turns a double play, or loads the bases. At the same time, I’m unhappy if the opposing team does the same thing.
I’m happy when my team is up by two runs going into the ninth inning, because it means they’re likely to win. I’m unhappy if they blow that same two-run lead, because it means they’re a lot less likely to win. I’m unhappy if my team goes into the ninth down by two runs, and I’m happy if they score a couple of runs to tie the game in the ninth.
All fans of every sport do this. If you don’t pay attention to your team’s chances of winning, then why watch at all? Because you like the uniforms?
Yet as soon as someone puts a number of it, you insist that it’s disconnected from the reality of what’s happening on the field. Why? WP is a reflection of what’s happening in the game as it’s being played.
